1. dataset

dataset：提供一种方式去获取数据及其 label

• 获取每一个数据及其 label
• 告诉我们总共有多少的数据

from torch.utils.data import Dataset
from PIL import Image
import os

class MyData(Dataset):
    # 初始化
    def __init__(self,root_dir,label_dir):
        self.root_dir = root_dir
        self.label_dir = label_dir
        # 路径拼接
        self.path = os.path.join(self.root_dir,self.label_dir)
        # Get a list of all pictures
        self.img_path = os.listdir(self.path)

    def __getitem__(self, idx):
        img_name = self.img_path[idx]
        img_item_path = os.path.join(self.root_dir,self.label_dir,img_name)
        img = Image.open(img_item_path)
        label = self.label_dir
        return img,label

    def __len__(self):
        return len(self.img_path)

root_dir = "dataset\\train"
blur_label_dir = 'blur'
sharp_label_dir = 'sharp'

blur_dataset = MyData(root_dir,blur_label_dir)
sharp_dataset = MyData(root_dir,sharp_label_dir)
img,label = blur_dataset[1]
img.show()

2. dataloader

dataloader：为后面的网络提供不同的数据形式（打包）

CLASStorch.utils.data.DataLoader(dataset, batch_size=1, shuffle=None, sampler=None, batch_sampler=None, num_workers=0, collate_fn=None, pin_memory=False, drop_last=False, timeout=0, worker_init_fn=None, multiprocessing_context=None, generator=None, *, prefetch_factor=2, persistent_workers=False, pin_memory_device='')

shuffle：是否打乱,false,不打乱（default: False）.
num_workers：How many processes are there when loading data（default: 0）.
drop_last：最后按 batch_size 取数据的话,See if there is any leftovers at the end.true,not enough leftbatch_sizedata will not be retrieved,false,The rest of the data will also be fetched（ default: False）.